#!/bin/bash
echo "=== STARTING AI SERVER (DUAL-CORE OPTIMIZED) ==="

# Bersihkan proses
pkill -9 ollama || true
rm -rf /tmp/ollama*

# Optimasi RAM & CPU
export OLLAMA_NUM_PARALLEL=1
export OLLAMA_MAX_LOADED_MODELS=1

# 1. Jalankan download model Python di background
python -B download_models.py &

# 2. Jalankan Ollama
ollama serve > ollama.log 2>&1 &

# 3. Tunggu API
for i in {1..60}; do
    if curl -s http://127.0.0.1:11434/api/tags > /dev/null; then
        echo "Ollama Ready"
        break
    fi
    sleep 2
done

# 4. Pull Models (DeepSeek & Qwen)
echo "Pulling DeepSeek-R1 7b (this may take a few minutes)..."
ollama pull deepseek-r1:7b

echo "Pulling Qwen-2.5 7b Instruct (Paling nurut format JSON)..."
ollama pull qwen2.5:7b-instruct

# 5. Run FastAPI
exec uvicorn app:app --host 0.0.0.0 --port 7860 --workers 1 --timeout-keep-alive 600